Up next in 10
Scale your CI/CD setup for Cloud Native Microservices via InnerSource by Turja Choudhary
2K views
Oct 30, 2023
C# Corner - Community of Software and Data Developers https://www.c-sharpcorner.com Conference Website: http://conference.c-sharpcorner.com #CSharpCorner #CSharpConf22 About C# Corner C# Corner, headquartered in Philadelphia, PA, is an online global community of 3 million software developers. C# Corner serves 5+ million visitors with 9 million page views each month. We publish the latest news and articles on cutting-edge software development topics. Developers share their knowledge and connect via content, forums, and chapters. Thousands of members benefit from our monthly events, webinars, and conferences. We also provide tools for career growth such as career advice, resume writing, training, certifications, books and whitepapers, and videos.
View Video Transcript
0:00
Let's start with our first poll
0:18
Are you guys asleep? Yeah. No, but yeah, I mean, at least some of you attended
0:24
and thank you for that. so again I just wanted to thank you the C-sharp corner the annual conference and to all the
0:35
organizers a big shout out to the organizers and also to the volunteers you guys are doing really
0:41
good work thank you thank you for organizing this thank you for inviting us over and a huge shout
0:46
out to the audience okay some of you who are here really thank you for attending I know this is like
0:53
a dead slot right everybody's sleepy including me i woke up at 4am today to take the flight here
1:00
perfect so uh very briefly uh who i am what i do i am currently an assistant director cloud practice
1:07
at ey ey is earnest and young some of you might have heard of it it's one of the big four firms
1:14
so when we say big four is like pwc ey then deloitte and kpmg okay so mostly they used to
1:21
initially do audit assurance tax but now they've also started doing consulting so i'm part of that
1:27
before this i used to lead the cloud center of excellence ccoe team in accenture so i have been
1:33
doing this for around 10 to 11 years now i started my journey with samsung research as an rnd engineer
1:40
then i went to pwc pricewaterhouse coopers uh uk then i came to accenture and now finally i am at
1:49
EY. Out of the 10 years, I have been doing around five years cloud computing. Okay, so
1:56
I have been with cloud for, started with AWS, now mostly in Azure. I hold around 40 plus
2:02
cloud or IT certifications and I'm also speaking at technology conference and I'm a DevOps
2:07
Institute ambassador. Okay, so a very quick poll, what do you guys do when I say the option
2:16
just raise your hand. College first and second years? Okay one, two, perfect a lot
2:23
of them, good thank you. College third and fourth years? Okay a lot of you, so yeah
2:30
most of you are from college including you. Working professionals? One to five
2:35
years? Okay one, one. Five to ten? Nobody, no issues. Ten to twenty? One, thank you for
2:44
joining us and 20 plus, oh 10 to 20 you are also perfect, thank you. 20 plus, oh nice, thank you so
2:51
much for joining us. Okay, so see don't think of this session as me telling you something, okay
2:58
think of me as a friend or as a colleague, just bouncing off some ideas with you, okay. And because
3:04
many of this session, much of this session was more focused on enterprise clients, okay, like
3:10
executive management and stuff but I see most of you guys are from college and
3:15
first to fourth years so I will skip some of the slides and focus on some of
3:19
the 101 slides more but in the interest of some of the working professionals who
3:24
attended I will kind of try to keep it balanced. So again this is a strategy
3:28
session okay so I will try to kind of invoke the thought about some of these
3:33
topics and if you can kind of even think five to six minutes on some of these
3:37
topics, that's good enough. Okay, you don't need to take home anything other than some of the
3:42
thoughts I'm talking to you about. So basically, let's discuss the art of the possible. Okay
3:48
so why I'm saying, let's just keep it at that. When we focus too much on tactical implementation
3:55
we sometimes forget about the strategic focus of vision, which is true. Okay, so for example
4:02
I mean, even for example, when you're giving an exam, maybe when you're just thinking about
4:06
for example, scoring the next marks or something. You might sometimes even forget, for example
4:10
what do you, how does it end up or how does it add up to the next phase of your life
4:14
What's maybe it's a job or master's degree or something like that. So let us today focus on
4:19
the strategic part of things. Some assumptions I have made, but again, I think most of these
4:25
assumptions are wrong. That most of you guys are aware of cloud computing. Maybe some of you are
4:30
at a high level, but I will try to go into that. And you have some basic idea about microservices
4:35
because that was the essence of this session. But again, I will have to kind of change my session accordingly
4:42
I will try to leave some questions, leave some time at the end for questions
4:45
But again, I am here for the next two to three days. You can kind of catch hold of me and ask me about anything
4:52
Okay, I've been doing this for around 10 years, so I've been in this industry for quite some time
4:56
Okay, let's get this party started, right? So scaling your CI-CD setup for cloud native microservices via inner source is a very complicated topic, right
5:07
I mean, so we will kind of try to remove the complications out of it. Again, this is an agenda, but I will not stick to this agenda because the audience has changed, very frankly
5:17
So let's just kind of think about microservices. So how many of you, for example, can somebody tell me what is a microservice? Anybody
5:26
Do anybody of you have any idea? Yeah? Exactly. I think, yeah, you will, I will obviously ask you, but anybody else in college has every, anybody heard about the term microservice
5:35
Okay, no, but no, have you heard, for example, okay, let me first start with you
5:41
Can you share something about microservices or something like that? Yeah? Okay
5:50
True. True. Okay. Yes, true. Okay, so let me ask to some of the people. I mean have any of you ever designed a web application
6:01
Okay, some of you have, right? Okay, you have. Perfect. So say for example, you have designed a web application. So what did it have
6:10
It had maybe some HTML5 pages, maybe as a front-end or something, maybe a little bit of JavaScript
6:16
And maybe what did you use for kind of like the backend coding, C-sharp, Java, something of those sorts maybe
6:22
maybe. So see the problem with enterprises is that they are doing what you are doing
6:27
at a very, very huge scale, which is the truth, right? They are doing at a huge, huge scale
6:31
For example, we are doing some of the activities where we are getting 1.89 million hits on
6:41
some of our systems every minute, which is like quite a big thing. So what, I mean logically
6:46
if you think about it, if there is a very big system, okay, it's very hard to manage
6:51
That's the truth. Anything in our life, if it's too big, it's very hard to manage
6:55
So what is our logical way of thinking about it? Say for example, you face a huge problem
7:00
you have to go to this friend's house, that friend's marriage and stuff like that. How do you kind of logically isolate stuff, right? Okay, let me plan my, say, Diwali in this way
7:09
Let me spend the first two days with my girlfriend or boyfriend. Let me spend the next three days with my parents
7:15
You try to break down a big problem into small, small problems, right
7:19
Because that is how we feel that, okay, I can control that
7:23
And I can kind of interconnect them. So this is microservices, basically. When you think of a very big system, okay
7:29
and you kind of split it into small, small subsections. And each of those sections are independently controllable
7:37
This is very important. I need to be able to control it. Why? Because if I can control it
7:42
I will be able to basically manage its scale. I will be able to ensure that, for example
7:47
if it goes down, the other part of my application doesn't go down
7:51
Say for example, if I tell you, say you have made a plan in Diwali with a single friend
7:56
five days you are going with him or her to Uti. Okay, and then for example, he or she falls sick
8:03
You don't have any backup, right? That is the idea. But if for example you had split that plan into say two and three days at least if two days would have been lost you would have still another three days So if a big problem you can make up into small problems similarly enterprises like say EY or any other enterprises we try to break big problems into
8:22
small problems. And this is microservices and very frankly this is kind of the talk of the town now
8:28
Maybe in a college level you might not be that exposed to it, but when you join, say for example
8:34
some of you are in third or fourth years, in one or two years you will join a company, right? Maybe
8:38
an IT company, then this will be extremely important for you guys. So as I said, microservice
8:43
architecture, it's like small, small services. You take a very big service, and I will show you
8:49
something along those lines. Say for example, if I take this. So this is for example, in the left
8:55
you have a very big monolithic kind of an architecture. What is a monolithic architecture
9:00
It's like a very big building. Everything is in there. And what is a microservice kind of an
9:05
architecture? You have split some parts out of it and you have made them independently scalable
9:11
independently deployable. So if you see here, I had a UI layer, I had a business layer
9:18
and I had a data access layer. These are the three components I had of my application. It's
9:22
a very standard web application. Most of the applications that you guys will work with in a
9:26
a few years, it will have a front-end, right? That is the one you interact with. And then
9:31
in the middle, you will have some sort of a compute. So say for example, as an user
9:35
I will go to the front-end, meaning I will say for example, there is an application to
9:40
upload images. Okay, so I will go, I will basically act as the, I will interact with
9:45
the front-end, I will give my data, I will upload my image, click on a button, but that's
9:50
my work, I don't do anything else. Then what happens basically, there is a middle where
9:54
something happening in the backend which processes this. So if you take both of this into a same
9:59
system, then your system is a monolith, so to speak. If you can break your front end separately
10:05
and your backend separately and deploy them independently, then it becomes kind of like a
10:10
microservices. Okay, so now coming to cloud. So I mean, in the last session also we were discussing
10:17
about cloud, right? So I mean most of you are aware of cloud at least, right? I mean you guys
10:22
have heard about cloud. Have any of you heard about mostly I'm referring to the college students
10:26
because most of you might have heard about who are working cloud native. Have any of you heard
10:31
about the term cloud native? No maybe. Okay so that's very interesting also. So have you heard
10:37
about the term native when we say I am native Bengali meaning I was born and brought up in
10:44
Kolkata I can speak the Bengali language similarly somebody is like for example a native Rajasthani
10:49
Okay, similarly so basically native means it's in my blood. So when you are doing something on cloud, okay
10:57
There are many ways you can do it. So for example, what is cloud as somebody was saying cloud is somebody else's computer, right
11:03
Somebody was saying it. So basically it's like I have a huge amount of compute. I
11:09
have a huge amount of storage So these are the most two important things that anybody needs to deploy an application. You need compute
11:17
So even your mobile phone, it has this 4v CPU and all of these things like Snapdragon
11:23
processor, Qualcomm processor, right? 6GB RAM and then you also need storage. That is where
11:28
we store our photos. If you have your compute and you have your storage, then you basically
11:35
have a system that you can work with. Now there are many ways to handle this in a cloud
11:40
environment. So what is one way? You for example, rent a virtual machine. Basically it's a server
11:46
What's a server? Server is a compute that you can work with. Now say for example in
11:51
your server you are deploying an application. That is very simple, you get a server, you
11:56
RDP into that server, you log into that server and you deploy your application. But is it
12:02
cloud native? No. That might not be cloud native because you are not taking the full
12:06
power of the cloud. So what is the full power of the cloud? The power of the cloud is that
12:12
I only pay for what I use. That is the most important power of the cloud. So say for example
12:19
today when you go to buy something, right? Say for example you are going to buy bread
12:26
you are going to buy bread, but you only need three breads basically. You just need three
12:30
breads but they are not selling you individual breads, right? So you get an entire bread box
12:35
kind of. So maybe you are paying 15 rupees whereas you only needed three so you should
12:39
have paid maybe 3 rupees. So this is what cloud is giving you. You don't have to buy
12:44
or rent a big server. You can basically just rent that server per minute, even per second
12:52
So say you start the server now, in one hour you stop the server, you just pay for that
12:57
hour. That is very important. That was not possible earlier. So this is cloud. Similarly
13:02
you have the power of independently scaling. So there is infinite, the concept of infinity
13:09
backed by infinite compute and infinite storage. So this is also the idea. Now what is cloud
13:14
native? Cloud native is application programming that takes full advantage of cloud computing
13:21
So your application needs to be deployed, needs to be created in a way that takes advantages
13:27
So our platform is providing you a lot of advantages, right? Say for example in your
13:31
house you have a microwave that's very useful for you to cook or heat food. But if you don't
13:37
take advantage of the micro when you are still using say for example a rovan or you are using
13:41
the same kind of attitude or same kind of practices then it will not work. So this is
13:46
what cloud native is about. So what is the problem today with enterprises? Again I am
13:53
trying to simplify this as much as possible because most of you guys are from college
13:58
So I will tell you what is the problem say for example say you guys are starting a party
14:03
five of you are starting a party and you buy alcohol as simple as that okay and you start a
14:08
party there are five people you buy say two bottles and a few glasses make it 50 50 people
14:17
still manageable okay I mean you can kind of call your friends get the alcohol get some more glasses
14:23
make it 500 now it's a little bit difficult where will I keep these people where will I get so much
14:29
of alcohol who will pay for it. Make it 5,000. Make it 50,000. Now make it 5 lakh. So a party
14:36
of 5 lakh people, where will I get the alcohol man? I don't know. Very frankly I don't know
14:41
right? So anything at scale is a problem. This is the problem that all enterprises will
14:47
hire you to solve. That is the single problem they are trying to fix. The same problem that
14:53
you have been able to fix in your day-to-day life, if you multiply it with a threshold of
14:58
say 1 lakh, 20 lakh, 50 lakh, 80 lakh. I don't know whether you guys remember when Flipkart was
15:04
doing that big billion sale, right? Every time in the initial days, the moment they would announce
15:10
that I will give you an admin, 2,000 rupees and stuff, people will log into the platform and the
15:14
system will crash. It's a very standard thing that used to happen in the beginning days. Why? Because
15:19
of the scale. Because of like say when 200,000 people are trying to log into the platform
15:24
it's a whole together different thing than you doing it on your laptop. So what can help you in scale
15:31
Can you tell me? What is the only solution to fixing problems at scale
15:37
Autoscaling is obviously a solution but I am saying from a generic concept. Say for example
15:45
if you think about that party concept. Okay so you were you are very okay with organizing a party
15:52
for 500 people, right? But when I ask you to organize a party of 500,000 people, it's
15:57
a problem, right? And how do you fix that problem? Standardization. What do I mean by
16:01
standardization? Put in a process. Ask everybody to bring their own liquor. Ask everybody to
16:07
bring their own glass. Right? Have a WhatsApp group. Everybody should log into that group
16:14
tell them yes, yes, yes, or maybe create a form or something. So basically any form of
16:19
standardization can help in fixing problems at scale And what is the biggest enemy of fixing problems at scale Non or exceptions There will be a friend who will
16:31
call you, you know, the alcohol shop is closed near my home, I cannot bring, please arrange
16:35
And that is how the problems will start, right? So anything that you cannot fix at scale is a
16:41
problem. And this is what we as enterprises try to solve every day. How can we, for example
16:47
fix problems at scale. Okay, perfect. So, many of these things I will kind of... Okay
16:53
So now from a standardization perspective, again, like I was trying to say, we have to
16:58
standardize things. So what are some of the things we can standardize? We can standardize
17:03
the architecture. What can I say? Say for example, I have started a company. Okay. I
17:07
have 20 people. Now, five developers are using JavaScript, five are using C-sharp, five are
17:13
using Java and five are using, say, Python. Everybody's coming to me, sir, this is not
17:19
working in Python. This is not working in JavaScript. What can I do? There is too much
17:26
movement, right? Too much chaos. So we need to standardize the programming language. There
17:31
are enterprises, for example, there are enterprises I've worked with which have told people, you
17:36
have to use C Sharp. There is no other way because it's a standardization. Again, we
17:42
don't want exceptions, we want standardization. So this is a very important concept that some
17:47
of these concepts if you can take with them today, that is even okay because many of these
17:52
technical aspects might not seem relevant to you today but at least some of these concepts
17:57
if you can kind of think about it and answer some of them it will be helpful. So architecture
18:02
programming practices and patterns, these are some of these things that are very interesting
18:08
and you have to standardize them. Everybody in your company should do it the same way
18:14
As simple as that. All the hotel rooms today have for example one say bathroom hypothetically
18:20
saying or two windows. You will not find one room with three windows for example. All the
18:25
rooms of the same category. Maybe a deluxe room might have three windows. So why did a
18:30
hotel build? Because it's very easy to scale. Everybody knows in every room you have to
18:33
there is no surprise. So always try to standardize stuff. This is what enterprises are looking
18:38
looking for. Okay, have any of you heard about CI CD? Okay, okay. Maybe you can share something
18:45
about CI CD? Continuous delivery, exactly. So what is... Incremental delivery. So incremental delivery is actually realized with DevOps. So they
19:04
are not conflicting, they are rather complementary. And I will discuss maybe separately with you
19:09
no worries. So they are actually complementary. Anyways, so what is CI CD? It's the fact of
19:15
building an application. So how do we build an application? Say for example, we have a C-sharp
19:19
code. You go to Visual Studio or Java, maybe Eclipse, you right click build. Right? What it
19:25
does, it generates a jar file or some sort of a package. That package is stored somewhere and that
19:29
That is called CI. And when you deploy that package to a kind of a compute environment, that's called CD
19:36
As simple as that. But what is the problem today that enterprises are facing
19:41
Enterprises are facing a problem called pipeline sprawl. This is what we are trying to say
19:47
When you have 4000 teams in your organization, each developing pipelines in a different way
19:55
it's a big problem. might not be a problem for small startups but for big enterprises it's a problem and what is the
20:01
solution? Standardization. So the same solution so we have to find a way to standardize our pipelines
20:09
Every single pipeline should follow a set of standard practices and patterns. They should not
20:15
be different. Obviously there will be nuances so my party will be a little different than your party
20:20
Maybe in my party I am allowing say for example water in your party you are allowing alcohol
20:24
hypothetically saying. But the idea is the same. So how do I standardize pipelines? And we have a
20:30
central kind of a team, again I'm moving a lot of this, called DevOps Dojo. So do you, anybody
20:36
have any of you heard about DevOps? DevOps you have heard, right? But I don't think anybody of
20:42
you have heard about DevOps Dojo. Right? Okay. So DevOps Dojo is another part of DevOps where we are
20:48
giving. So what is Dojo? Dojo if you search, it's like in Japanese, people are kind of doing Kung Fu
20:53
and karate in a ground. Okay, it's like a dojo. They call themselves dojo and there is a master
20:58
and people. So Bruce Lee movies, they go to a dojo and they fight amongst themselves and stuff like
21:03
that. Okay, so DevOps dojo is a similar concept of very highly experienced DevOps people. Okay
21:09
because I am part of the DevOps dojo team. So we are a very highly experienced DevOps people. We are
21:15
kind of creating those basic pipelines, the basic standards that we are then asking all the other
21:21
teams to follow. Again, what is the advantage? Everything is standardized. Okay. So this is a
21:27
reference application or a reference. For example, I have my DevOps Dojo team. So I am part of the
21:32
team. I have my security team. I have my DevOps engineers. All of us are brainstorming and creating
21:38
a standard pipeline. So that standard pipeline I'm storing in a repository. Then what I'm saying
21:44
anybody needs, say for example, I have created this standard kind of an architecture standard
21:48
Now say for example you need something else. You are a business team, you want to deploy
21:52
e-commerce application. Okay, you don't have to develop everything from scratch. You come
21:58
to my team, you ask me, give me the standard. You will take the standard, you will download
22:03
it and you will modify some of the parameters and you will have a working pipeline basically
22:08
Basically that is the idea. Yeah. Yeah, so DevOps governance model, if you see like a DevOps co-op operating model, DevOps model
22:26
DevOps Dojo is a part of the architecture governance board. But it's not just the only part
22:30
So it's not like that they are leading. They are a part just like many other parts. For example
22:35
you might have representation from other teams also. Okay. So anyways, let me now, let me very
22:40
quickly come to something. How many of you have heard about open source? Perfect. How many of you
22:46
have heard about inner source? Okay. So who has heard about open source? Again, anybody? Okay
22:52
man, can you just help me out? What is open source to you? Open, exactly. Exactly. So basically
23:01
it's like the code is in GitHub as simple as that right. So for example, I am able to go to GitHub
23:07
download the code, build the code, test the code, make some changes, and then again upload to GitHub
23:13
I can contribute basically, right. So say for example, this is an open source community, right
23:18
We call them an open source community. So there is people say for example, you have created a
23:23
code in India in Delhi, and maybe there is an issue in your code and somebody in Brazil is fixing it
23:29
it. That happens in open source, right? Because everything is open. You have very good documentation
23:34
you have very good practices. Anybody in Brazil can go to your code, go to your report, download
23:38
it, okay this is a problem. Fix it, raise a pull request. You are a reviewer, you merge
23:43
it. Very good. This is the basic fundamental of open source. Complete collaboration, complete
23:49
transparency. No working in silo. Everybody knows everything, right? So inner source is
23:56
open source in an enterprise. So now think of open source within say for example a company
24:02
like say Accenture. Everybody is doing their own thing. There are different teams within
24:07
Accenture. But my team has no idea about your team. Say for example I have created something
24:12
a web application and you have created, I have created the front end, you have created the back end. My front end is calling your back end. Okay. But your back end has an issue
24:21
is my option? I have to talk to you. I cannot go into your code and fix it. I cannot today
24:27
If you see our siloed model inner source breaks that In inner source the concept is very simple So if I go here say for example you have the main branch where the actual code is residing Okay that is the main branch But you for example you
24:41
are curious enough and you are knowledgeable enough that you want to fix some issue. You
24:46
just create a branch from the main branch, just like open source, you commit the code
24:52
and you raise a pull request, it goes to the main. So basically open source concept, if
24:57
If you take within an enterprise, it's called inner source. I'm simplifying a bit, but that's it basically
25:03
So how can it help? How can inner source help? Inner source can help in the same way
25:08
that open source helps people. Say for example, I am a team of five people
25:12
just five people, I have created a web service, right? Say for example, there are many issues with the web service
25:18
I am only five people, we can only fix as many issues, right? But say for example, I have very good documentation
25:24
I have a very good inner source kind of operating model. The entire company becomes my workforce, theoretically
25:32
Anybody in my company can go into my code, find that issue, and fix it
25:38
That is the concept of inner source. And again, that is what I'm trying to say
25:42
What was the problem that we were trying to fix? We are trying to scale things
25:46
We are trying to handle teams and handle issues of 400,000 developers
25:52
It's not one developer, two developer. It's like 4,000 teams, 400,000 developers
25:57
So the only way to fix it is with scale. So I as a team of five members cannot fix this, right
26:03
So I need the entire company behind me. So that is what we are now trying to do
26:07
We are trying to increase the adoption of the microservices by having pipelines, which are standard pipelines
26:15
and then we are asking different teams to contribute to those pipelines
26:20
with an inert source approach. Again, I can understand there is some confusion because again these are a little bit complicated
26:27
topics, okay, more suitable for enterprise audiences, but I'm trying to make it as simple
26:34
as possible for college students. So what is the key learning from all of this
26:38
Very quickly, right? So if for example we kind of summarize what I was trying to say, anything that you do at
26:44
scale is an issue. Whether it's organizing a party of 500,000 people, or is it a company like EY with 3,50,000
26:51
developers. Okay, so anything at scale is an issue. And what is the way to solve problems
26:58
at scale? Standardization. So wherever possible, we should try to standardize things as much
27:06
as possible, whether it is standard code, whether it is following a standard architecture
27:11
whether it is following a standard programming language. So we should, for example, ask everybody
27:15
to shift to Java. Just because we will have all the resources, everybody, only in Java
27:19
will help our productivity in a sense. So all of this standardization you need, so
27:24
for a microservice and a CI CD try to create CI CD standard pipelines that can
27:30
help you and then have inner source methodologies and processes so that the
27:35
entire company becomes part of your team. It's not just five people or ten people
27:40
who are working on a project with a good inner source operating model. The entire
27:45
company is basically part of your workforce. Okay but anyways I will try to
27:49
to kind of... Yes, sorry. Yeah, yeah. No, no, no. See, enterprise asset libraries are always
27:59
there kind of, right? Many companies had standard libraries. So for example, for example, there
28:04
were companies where, say you want to do a data load. Say there is a file, you want to
28:09
load that file into some sort of a data store. There were standard ingestion methods. Everybody
28:15
was using it. But there was only one team who used to maintain that. Now we are going
28:20
a step further and we are saying you can also raise a pull request and change that. Exactly
28:26
Basically everybody. So it's a distributed, federated kind of a model. Okay? Perfect
28:33
I think. And let me just again, I mean just for your curiosity's sake, again, many of
28:37
these things will not maybe make a lot of sense to you because you have not again been
28:41
the IT industry when you start it will but the idea is only way to scale is as I told
28:48
you to have standard things and then make them easily accessible. So for example I need
28:53
a web application. You will see 80% of the web applications are the same. It is a front
28:58
end, it is a back end, it is a middleware. So why should I not be able as a user in a
29:02
company, why should I not be able to go to some sort of a marketplace just like Amazon
29:07
marketplace, I should go somewhere and select a web app. I will just click, I need a web
29:12
app, and a web app will be created for me. So that is the future. That is what we are calling self-service via IT catalog. And we want to have a cross-functional architecture
29:21
board. Again, this might not be very relevant. And again, implement inner source patterns
29:25
This is also very important. So with this, I'm kind of trying to keep it short. And is
29:31
there any question, for example, in any other topic, even for example, not maybe related
29:36
to this, maybe also any confusion with cloud, DevOps, any concept of DevOps, CI, CD, anything
29:42
that is interesting to you guys. Otherwise, I can end the session here
29:46
Yes, sorry. Yeah, so see, DevOps, so see, what is basically, what is a virtual machine, right? So virtual
29:58
machine is basically a physical machine which has been virtualized for you, and then you
30:02
are logging into that machine you get say 4 vcpu and 128 gb ram basically okay and you
30:08
have some data and you do it. Docker is basically you have a container run time on top of the
30:14
virtual machine so that you can run multiple container processes basically you take container
30:20
and so basically it's more lightweight than VMs. So how much time does a VM take to boot
30:24
up like I think 2 to 3 minutes whereas a container might boot up in a few seconds. Okay this
30:28
one it's stateless. So there I mean, how does it relate to DevOps? These are again, these
30:33
are not complement, contradictory stuff. These are all complementary stuff as part of the
30:38
overall cloud native model. So what is an overall cloud native model? You have basically
30:43
microservices. So what I'm, so say for example, you have a code. So where do we start with
30:48
We start with a code. We have written a code. We need to build that code into a Docker image
30:55
might have heard about a docker image. So how do I go from code to image? That is CI, continuous
30:59
integration. Simple as that. I have in my left hand source code. I apply a CI pipeline on top of my
31:06
source code. I get a container image. As simple as that. That is where my CI ends. I keep that
31:11
container image in a very safe, secure container registry. I have to keep the image somewhere
31:16
right? I keep it in an Azure container registry, hypothetically saying. So that is where my CI ends
31:20
But what is an image good? It has no value in itself. I have to deploy it to a compute to be able to derive some value out of it
31:28
So I take that image, I run a deployment pipeline called CD pipeline on top of it and I deploy it onto a Kubernetes cluster
31:36
Or to a Docker swarm cluster, any sort of a cluster. Why do we need a cluster? Because individual machines or nodes can go down
31:45
Every time, always remember, whenever we deploy something, we should have two instances of it
31:49
If one instance goes down, there will be another. So that is why we call it a cluster
31:53
Minimum three kind of a node to maintain the quorum. So this is a concept. They are not contradictory
31:58
Docker and Kubernetes are compute orchestration methodologies to deploy. CICD pipelines are automation or automated ways for you to help from source to destination
32:11
Nothing else. Okay, perfect. I think I am out of time also
32:15
Perfect. Thank you so much. and guys I mean if you want to connect with me or something like that
32:20
you can either connect with me later or you can connect with me
32:25
on LinkedIn also whatever you prefer but I have reached my maximum
32:31
connection so you will have to kind of follow me, okay, perfect
32:35
thank you, take care oh sorry sorry sorry Thank you
#Enterprise Technology
#Programming